From Big Data to Social Media Analytics

Introduction (1h)

  • Why now?
  • What is Big Data? volume, velocity, variety, veracity, … , and value
  • Paradigm shifts enabled
  • Market Landscape

Mastering the Volume Dimension (9h)

Operational perspective: the NoSQL world (3h)
  • Recalling basic concepts of the relational model
  • NoSQL: basic concepts
  • Key-Value stores
  • Column-family stores
  • Hands-on HBase
  • Document-based stores
  • Hands-on MongoDB (?)
  • Graph-based stores
Analytical perspective: from Map Reduce (hadoop) … (2h)
  • fundamentals
  • pros and cons
  • evolution
  • eco-system
Analytical perspective: … to Spark (4h)
  • Introduction
  • Hands-on: simple Apps
  • Essentials
  • Hands-on Spark SQL
  • Hands-on MLlib, k-means
  • Hands-on GraphX

Mastering the Variety Dimension (1h30m)

  • Variety is unavoidable
  • Embrace variety with semantic technologies
  • Demonstration of ontop

Mastering the Velocity Dimension and beyond (2h30m)

  • It’s a streaming world
  • Information flow processing
  • Hands-on Event Processing Language
  • Volume+Velocity: Hands-on Spark Streaming
  • Velocity+Variety: Stream Reasoning

The “Traditional” volume-centric use cases and case studies (2 hours)

  • The Forrester WaveTM to choose among the vendors
  • Amazon
  • Cloudera
  • Hortonworks
  • MapR technologies
  • IBM Big Insights
  • Microsoft Azure HDInsight

Social Media Analytics (5 hours)

  • what’s social media?
  • why shuould I care?
  • Personal Social Media Analytics
  • Social Media Analytics for companies
  • So, where’s Big Data?
  • From micropost to (Big) data
    • Collecting Social Media
    • Named Entity Recognition
    • Entity Linking
    • Sentiment/Opinion extraction
  • Example of Social Media Analytics

The 5 game changing big data use cases (6 hours)

  • Introduction to the 2015 study by IBM Analytics on the five high-value use cases that can be the first step into big data
    • the study
    • how to read it
  • Use cases solved with IBM technology as well as with other products offered on the big data market
    • Big data exploration use case
    • Enhanced 360-degree view of the customer use case
    • Security/intelligence extension use case
    • Operations analysis use case
    • Data warehouse modernization use case
  • Conclusions

City Sensing case study (2 hours)

  • The digital reflection of our cities is sharpening and it is tracking their evolution with a decreasing delay. This happens thanks to the pervasive deployment of sensors, the wide adoption of smart phones, the usage of (location-based) social networks and the availability of datasets about urban environment.
  • So while data becomes every day more abundant, decision makers face the challenge to increase their capability to create value out of the analysis of this data.
  • This part of the course presents how advance visual analytics, ontology base data access and information flow processing methods can help in making sense of Social Media Streams and Call Data Records from Mobile Network Operators during city scale events. Real-world deployments demonstrate the ability of those methods to advance our ability to feel the pulse of our cities in order to deliver innovative services.

Spark case studies (1 hour)

  • Spark at Twitter
  • Hadoop and Spark Join Forces inYahoo
  • Collaborative Filtering with Spark at Spotify
  • Stratio Streaming: a new approach to Spark Streaming!
  • Sharethrough Uses Spark Streaming to Optimize Bidding in Real Time
  • Guavus Embeds Apache Spark into its Operational Intelligence Platform Deployed at the World’s Largest Telcos
  • One platform for all at Conviva: real-time, near-real-time, and offline video analytics on Spark
  • others from Spark summit 2015